[Rd] call to pt: Warning message: full precision was not achieved in 'pnt' (PR#9171)

FromeEL at ornl.gov FromeEL at ornl.gov
Thu Aug 24 01:58:29 CEST 2006


I have followed all instruction in R FAQ Section 9.2
This looks like a bug to me

"C:\Program Files\R\R-2.3.1\bin\Rgui.exe" --sdi
> version
               _
platform       i386-pc-mingw32
arch           i386
os             mingw32
system         i386, mingw32
status
major          2
minor          3.1
year           2006
month          06
day            01
svn rev        38247
language       R
version.string Version 2.3.1 (2006-06-01)
>
# If a command you are familiar with causes an R error message in a
# case where its usual definition ought to be reasonable, it is
# probably a bug.

     call to pt with ncp < 10^-5
     Warning message:
     full precision was not achieved in 'pnt'
#
# example
> nc
[1] 1e-01 1e-02 1e-03 1e-04 1e-05 1e-06 1e-07 0e+00
> cnpt
function(p,df,nc){ pt( qt(p,df),df,nc) }
> cnpt(.95,10,nc)
[1] 0.9395967 0.9490285 0.9499035 0.9499904 0.9499990 0.9499999 0.9500000
[8] 0.9500000
Warning messages:
1: full precision was not achieved in 'pnt'
2: full precision was not achieved in 'pnt'
3: full precision was not achieved in 'pnt'
> sessionInfo()
Version 2.3.1 (2006-06-01)
i386-pc-mingw32

attached base packages:
[1] "methods"   "stats"     "graphics"  "grDevices" "utils"     "datasets"
[7] "base"

>#   quit 2.3.1
>q()


### run 2.2.1
"C:\Program Files\R\R-2.2.1\bin\Rgui.exe"  --sdi

> version
         _
platform i386-pc-mingw32
arch     i386
os       mingw32
system   i386, mingw32
status
major    2
minor    2.1
year     2005
month    12
day      20
svn rev  36812
language R

> cnpt(.95,10,nc)
[1] 0.9395967 0.9490285 0.9499035 0.9499904 0.9499990 0.9499999 0.9500000
[8] 0.9500000
>
#  no warning message in previous versions of R

=========================================================================

Thanks for your advise on this issue

Ed Frome

http://www.csm.ornl.gov/~frome/

	[[alternative HTML version deleted]]



More information about the R-devel mailing list